Caterpillar (CAT) vs Copper

Caterpillar's revenue is roughly 50 percent Construction Industries, 25 percent Resource Industries (mining equipment), 20 percent Energy and Transportation, and 5 percent Financial Products. Copper is the most economically sensitive industrial metal, with end-use demand split roughly 28 percent construction, 19 percent electrical grid, 17 percent transportation (with a rapidly growing EV component), 16 percent consumer goods, 11 percent industrial machinery, and 9 percent other. Both series are dominated by China demand at the marginal-buyer level: China consumes approximately 55 percent of global refined copper and is Caterpillar's second-largest equipment market. The pair has been the headline cross-asset cycle bet for over two decades, cited in BHP's Investor Briefings, Glencore's commentary, the CME Group's copper supply-and-demand reports, and the Federal Reserve Bank of Kansas City's industrial-cycle research. The institutional thesis the pair tests is whether equipment demand (CAT, the user) is leading or lagging the copper cycle (PCOPPUSDM, the input). The recurring breakpoint is the 2022 IRA-era rerating, when copper demand-growth narratives re-anchored on EV and grid electrification rather than China construction. The composition of the marginal copper buyer therefore shifted, and the historical CAT-copper correlation that ran above 0.85 in 2014-2016 has not been recovered since the 2022 break, even though the rolling correlation has stabilised around 0.71 in 2025-2026. The composition shift matters because each marginal buyer has different price elasticity and different volume sensitivity, and a copper price driven by EV and grid demand carries different forward-return implications than the same price driven by Chinese construction.

The CAT-versus-LME-copper rolling 12-month correlation has run 0.62 to 0.78 across 2010-2026, with three regime breaks worth understanding. First, 2015-2016 China commodity slowdown: copper fell from $7,400/t in mid-2014 to $4,400/t in early 2016 (a 41 percent decline), and CAT fell from $112 to $59 (a 47 percent decline) over the same window, with the 12-month correlation peaking at 0.91. The episode is the cleanest case for treating the pair as a unified cycle bet, and worth a more granular reconstruction: Chinese fixed-asset investment growth slowed from 19.6 percent year-on-year in mid-2013 to 9.2 percent by year-end 2015, the steepest deceleration of the post-2000 era, and both legs bottomed within four weeks of each other in early 2016. Second, 2020 COVID divergence: copper rallied from $4,617/t (March 19, 2020 trough) to $9,617/t (May 2021), a 108 percent advance, while CAT rallied from $90 (March 23, 2020 trough) to $244 (May 2021), a 171 percent advance. CAT outperformed copper because the equity carried the equity-multiple expansion that the commodity could not, but the rolling correlation stayed above 0.65. Third, 2022 IRA-era rerating: from August 2022 (Inflation Reduction Act passage), copper began to decouple from China-construction narratives and re-anchor on EV and grid electrification demand. Copper traded to $11,200/t in May 2024 and to fresh records above $11,200 in late October 2025. CAT followed but at a more measured pace, with the rolling correlation falling to 0.58 in 2023 before recovering to 0.71 by late 2025. The lesson from each regime break is that the pair operates differently in supply-driven versus demand-driven cycle phases, and the correlation tightness is the most reliable indicator of which regime the market is currently in.

Three transmission channels link CAT to copper. The shared cycle channel: both series respond to global industrial production indices (JPMorgan Global Manufacturing PMI is the cleanest single read), with CAT typically responding faster on the equity side because earnings expectations re-rate more quickly than commodity inventory cycles clear. The marginal-Chinese-demand channel: both series moved together through the 2010-2022 era because Chinese fixed-asset investment was the marginal driver for both copper consumption and Caterpillar mining equipment orders. After 2022, this channel weakened as Chinese growth slowed sharply even as copper prices headed to new record highs. The discount-rate channel: CAT, as an equity, has higher implied duration than the copper price; rate-cycle moves affect CAT through the equity discount-rate channel and copper through the inventory-financing-cost channel, with the equity moving faster. The 2022 episode is the cleanest example: as the 10-year yield rose from 1.51 percent to 4.25 percent, CAT outperformed copper (positive 17 percent versus copper's positive 4 percent for 2022) because the equity discount-rate channel anchored to terminal-rate expectations while copper held against the China demand reset. Reading the pair through these three channels separately is the structural method, rather than treating it as a single shared-cycle bet. Each channel carries different forward-return implications and different lead-lag dynamics, and disentangling them is what separates a clean cyclical signal from a noisy one when the pair diverges from its historical correlation band.

Caterpillar full-year 2025 sales and revenues: $67.6 billion, up 4 percent from $64.8 billion in 2024. Adjusted profit per share 2025: $19.06 (versus $21.90 in 2024). The earnings decline reflected weaker construction-industries pricing combined with margin pressure from the Q1 2025 Service Industries weakness. CAT closed near $361 on April 29, 2026. LME copper near $9,800 per tonne in April 2026, after touching record highs above $11,200 in late October 2025 (LME forward curve hit those levels on supply-side disruption news, surpassing the prior record of $11,104.5/t set in May 2024). The copper-CAT divergence in 2025 has been the most pronounced since the 2020-2021 episode: copper at multi-decade highs while CAT lagged the broader equity market through the first three quarters of 2025. The mechanism is the shared-cycle channel weakening (Chinese construction slowing) combined with the EV/grid-electrification channel benefiting copper specifically rather than equipment demand. The April 2026 configuration is consistent with a transition phase: copper anchored on supply tightness, CAT working through the global industrial mid-cycle slowdown, and the next 6-9 months likely hinging on whether the Trump-administration infrastructure and reshoring agenda accelerates the equipment cycle in line with the copper price. The Q4 2025 earnings call flagged Resource Industries order book growth at 5.8 percent year-on-year, below the 12 percent threshold that historically signals an equipment-demand acceleration, and Construction Industries orders were down 3 percent year-on-year, both consistent with the transition-phase classification rather than a pre-acceleration setup.

The Convex Net Liquidity Impulse (CNLI) reads CAT and copper as separate inputs into the cyclical-growth basket. CAT carries weight inside the equity-cyclical bucket (alongside Deere, Cummins, and the broader S&P industrials index), and copper sits in the commodities-cyclical bucket (alongside iron ore, oil and the CRB Raw Industrials index). The pair is most informative when the two buckets disagree, which is the current 2025-2026 configuration: copper signals supply-tight cycle continuation, CAT signals demand-led mid-cycle slowdown. CNLI flags this disagreement as a cross-asset signal rather than a directional risk-on or risk-off classification, because the two buckets together tell a more nuanced story than either alone. The horizon for the signal is 6-12 months, with the historical base rate suggesting that copper-CAT disagreements resolve through copper compression rather than CAT rerating in the typical case (the 2014-2016 episode being the canonical example, where copper compressed faster than CAT despite both falling). The April 2026 configuration suggests the Trump-era infrastructure agenda will be the dominant variable for whether this episode follows that template or breaks the historical base rate. The CNLI cross-check is therefore the cleanest single read on whether the disagreement is informative or noise. The base-rate calculation is conditioned on the 2010-2025 sample, which spans two full cycles and is large enough for the 6-12 month forward distribution to carry meaningful statistical weight rather than reflecting a single-episode template.

Three indicators trip the pair from its current divergence regime. First, a sustained global manufacturing PMI reading above 51 (currently 49.8) would signal the demand cycle has turned and would favour CAT over copper, because CAT carries the operating leverage to a manufacturing-cycle upturn. Second, a copper LME price below $8,500/t would signal the supply-tight regime has resolved and would re-correlate the pair on the downside. Third, a sustained reshoring-driven Caterpillar Resource Industries order book above 12 percent year-on-year (it has run 4-7 percent through 2025) would signal that the equipment-demand cycle is responding to the same electrification and reshoring drivers that copper is pricing. The April 2026 configuration is in transition: rolling correlation 0.71 is above the 2010-2026 lower bound of 0.58 but below the 0.91 peak of 2015-2016. The pair is most actionable at the regime inflection points (2014-2016 turn, 2020 COVID, 2022 IRA rerating), and Convex publishes the regime classification daily on the cross-asset dashboard alongside the rolling correlation, the bucket-disagreement read, and the CNLI cyclical-growth composite reading. The data cadence is daily for both legs, which makes intraday divergences tradeable when the underlying regime classification is stable. The release calendar provides additional structural anchors: Caterpillar reports earnings quarterly (typically late January, April, July and October), and major LME copper supply-disruption news flow is irregular but concentrated around the Q4 mining-industry conference cycle, both of which create predictable pair-volatility windows that can be sized into trade construction.